For the record, im just sharing some observations and thoughts here. Im not claiming anything as fact. Nor is my purpose to point at anyone and saying their numbers are wrong. For those who feel affected, dot take this personal. I just want to get a discussion around this topic. If it can help measurements shared by BLF users to be more similar then nothing would be better than that.
The fact that the similar lights have slightly different readings, and that manufacturers who uses ANSI standards for measurements are all over the place does not help either getting everyone one the same page. But I think we, the enthusiasts should give it a shot.
The purpose of sharing measurements
What is the purpose of sharing measurements? Isnt it to be able to post numbers that are comparable with others? If so, would it not be beneficial that everybody were more on the same page? Would it not be beneficial for everybody to try and get closer to each other?
Lux:
When it comes to lux/kcd numbers its been quite clear for some time that the people with LX1330B have the highest numbers. At least that is what I have noticed. These numbers seems to some extent to have become close to the benchmark simply because so many of the people who often posts lux readings use one.
Typical difference in lux between a low reading meter and a LX1330B can be up towards 30%. In general though, it seems like peoples lux numbers are fairly close. Usually within 10-15%. Although, LX1330B usually have the peak numbers. That number is usually higher then most manufacturer ratings, and usually higher than calibrated expensive light meters.
Lumen:
There is mostly two camps here among people that often share numbers.
rdrfronty, Tom E, DBCstm, RMM and maybe others are example of people in the "high camp". Expect up towards 22% higher readings than most others. These people are among the main contributors of sharing lumen numbers on BLF. Most are into modified lights, so accuracy towards ANSI rated lights does not seem that important. Measuring 1200 lumen OTF from an emitter driven at 2,9-3A is also possible with the "high lumen calibration". Even though that is a bit doubtful compared to what you might expect from a Cree XM-L2 according to specifications. Thanks to copper mcpcb, and maybe extremely small OTF loss, it may be possible for all I know, but it certainly seems a bit too high for me.
Then there is the "ANSI camp". For these people getting a nice overall calibration compared to premium lights seems to have been of high priority and the baseline of calibration. Selfbuilt is one of the more known reviewers on CPF, he have used a lot of time calibrating his equipment for correct lumen readings. His numbers are basically the same as Fenix. Which is one of the manufacturers who posts numbers that seems quite consistent and reasonable (based on my personal observation and several others). I believe that some of the "top" reviewers and light buyers on BLF are typically within 7% of those numbers. _the_, johnnymac, HKJ, etc. I personally consider the work all these people have done with several hundreds of lights combined to be the benchmark. Usually when I see someone from another forum, theirs numbers usually are closer to the "ANSI camp". I dont see why this should not be the benchmark for everyone.
Some example issues:
People have a tendency to take numbers for a fact, or to be quite similar. Many are not aware that there can be up towards 30% difference in the numbers, and that 20% difference isnt that uncommon. This leads to issues when people want to compare different lights measured by different people.
In general, in some places its looked down upon to have too high numbers. Its often looked upon as cheating or bragging. On BLF, high numbers (posted by members) are usually met with "WOW, that is amazing. Great mod, etc). The higher the calibration, the more "wow". So if anyone wants to compare numbers with the people with the top numbers, you have to calibrate your gear for higher numbers. If someone with a "high calibration" sells lights, for many it will seem as if the modded lights are better then others, but really, the difference is just the calibration. At least that is what we have to assume, because the numbers are rarely comparable.
Its quite typical that someone compares manufacturer numbers. Say a TK61 with a modified TK61. A TK61 are rated at 1000 lumen. A modified might be measured to 1600 lumen. Many expect a 600 lumen/ 60% increase from stock. But they may not be aware that the guy who modded a light would measure the stock light to be 1200 lumen. Only making the increase 400lumen/33%. Quite a difference between a 33% gain and a 60% gain just because of a 20% difference in stock calibration.
Ive often heard about people who are not able to replicate numbers from other peoples mods. This may be due to issues with the mod, but in many cases, its just the calibration of the measuring equipment.
More example differences:
Here are some numbers for you. Please study them a little bit. All on Supbeam K50. There are many lights I could do a lot of funny numbers with. But K50 is one of the lights where many have shared numbers, and its fairly recent.
1600 Lumen - 140 kcd - Official numbers by Supbeam (Notice the high lumen number and "super low kcd")
1306 Lumen - 545kcd - JMpaul320 numbers from K50 modded vinh de-domed emitter and current increase (reference and more numbers here)
1450 Lumen - 188kcd - Selfbuilt numbers (Measured with calibrated lux meter, and the "typical ANSI lumen calibration")
1656 Lumen - 208kcd - (one of Tom E`s K50s. High lumen club + LX1330B light meter)
1642 Lumen - 218kcd - DBCstm stock K50 numbers. High lumen club + LX1330B light meter I assume. Quite similar numbers to Tom E, but even higher kcd)
Looking at JMpaul320 lumen numbers its obvious he belongs in the "ANSI rated lumen club". Despite the current increase, with losses from a de-domed emitter he still have the lowest lumen output, and its the only modified light. He probably uses LX1330B light meter. At least that is the meter he used before. His lux readings are always high, lumen readings seems to be typical ANSI numbers, although I have not studied them much.
If you ask DBCstm how many lumens his Supfire M6 with a FET driver have, he will say close to 5000 lumen.
If you ask JMpaul320 how many lumens his Securitying with FET driver (modified by RMM?), he will probably say about 3800 lumen. Sure, Dale`s M6 will have slightly more lumens with lower CRI U2 emitters and it probably have slightly less heat sag due to M6s better construction. But, it just shows to some extend the difference you can get by asking two people on pretty much the same light how the output difference is.
A possible improvement that I believe would benefit everyone:
The "high lumen club" adjusts down their lumen ratings with 7-15% (depending on how high their readings are compared to others.)
People with LX1330B (and other very high reading lux meters) adjusts down their lux readings with 2-7%.
Some last thoughts:
Sharing amp numbers, before/after mod, comparison with other lights always helps put numbers in a context. But in many cases, that is not done.
I think the purpose of sharing numbers is to contribute with data the is meaningful and helpful to others. If the numbers between different people have too large differences, it can be more misguiding than helpful. I believe that is often the case when "newcomers" check out numbers from "veterans" and assume they to some degree have comparable numbers and that these numbers can be compared with manufacturers of premium lights. Personally, I always have to "adjust" various peoples numbers in order to be able to get a better comparison. If the people with the "inflated numbers" would try and get on level with the "ANSI guys", then I think it would benefit everyone and make numbers more comparable and useful.
Wouldn't it be better if everyone tried to be on the same page?